Working with Images

[Stuart Rogers]

John Sgammato wrote:
> When you capture a 96dpi image at higher resolution, you will never
> see detail that isn't there (of course) but you can do more with the
> image because your OWN image of the image is capable of showing
> greater resolution. You can look at it as if your high-res image
> capture dices the existing image into smaller pieces. As an extreme
> example, consider an original image of alternating 1-inch black and
> white elements along a line at 10 dpi. Capture that image at 100 dpi
> and you really have 10 times as many 0.1-inch elements to work with,
> all faithful in location, dimension, and color to the original. If
> you need to rotate or stretch or manipulate the image in any way, or
> if any of your processes cause the image to lose resolution, the new
> hi-res image will be more forgiving. Likewise if you print the image,
> the printer is limited by its own resolution - the higher-resolution
> image can help to compensate.
> 
> This is easy to test for your self: in Illustrator (or similar)
> generate a black square and inside it a white circle or diamond.
> Repeat at smaller intervals until you get bored. Save as .ai, then
> export to .tiff twice. For the first select 96dpi and call it
> lo-res.tiff, and for the second export at 400dpi and call it
> hi-res.tiff. Then import them side-by side into FM and see how they
> look. The lo-res image will show jaggy edges that you don't see in
> the hi-res.

Hi John,

Sorry, but that's not how it works!

All that happens in a screen capture is that the capturing software 
copies the contents of all or part of the graphics card RAM to a file. 
"Resolution" is irrelevant at that stage, because you are only copying a 
fixed number of pixels.  Those pixels are displayed by your monitor 
according to the graphics card resolution setting, which determines the 
image dimensions *on your particular screen*, and they are (later) sent 
to a printer driver with an instruction on how closely to space the 
corresponding ink dots.  But none of that changes the number of pixels 
in either the graphics card RAM or the resulting file.

Also, your test doesn't apply to screen captures.
Illustrator is a vector program, not a raster program.  When you
export the vector drawings to tiff, they get rasterized (converted from
mathematical formulas with no associated quantity of pixels to files 
containing a finite number of pixels).  If you export at low resolution, 
then Illustrator will create a file with fewer pixels than if you export 
at higher resolution.  This export operation is completely different 
from a screen capture, which is a raster image with a fixed number of 
pixels.

"Jaggies" are unavoidable when rectangular pixels are used to create 
angled lines.  They're just less visible with higher-res files, though 
they are still there.

HTH!

-- 
Stuart Rogers
Technical Communicator
Phoenix Geophysics Limited
Toronto, ON, Canada
+1 (416) 491-7340 x 325

srogers phoenix-geophysics com

If it makes things work more easily, why isn't it called lubrican?